How Do Microtubules Grow and Shrink: Dynamic Instability

Microtubules grow by adding GTP-bound tubulin dimers to their plus end, and they shrink when GTP hydrolysis strips away the stabilizing cap at that tip, triggering rapid depolymerization. The whole cycle, called dynamic instability, flips between these two states stochastically, and a surprisingly thin layer of GTP-tubulin at the very tip is all that stands between a growing microtubule and a catastrophic collapse.

What a microtubule actually is

Think of a microtubule as a hollow tube built from stacked rings of protein. Each ring is made of 13 protofilaments, and each protofilament is a chain of alpha/beta-tubulin dimers stacked end-to-end. Because every dimer always orients the same way, the tube is polar: one end exposes the beta-tubulin face (the plus end) and the other exposes the alpha-tubulin face (the minus end). That polarity matters enormously because the two ends behave very differently. The plus end grows and shrinks fast. The minus end does the same thing, but much more slowly. In cells, the minus end is often anchored or capped by proteins like the CAMSAP/Patronin family, so most of the interesting growth and shrinkage action happens at the plus end.

The defining property of microtubules is dynamic instability: a single microtubule can switch abruptly from steady growth to rapid shrinkage (catastrophe) and then, sometimes, switch back to growth again (rescue). This is not gradual thinning. Shrinkage during catastrophe is about 10 to 100 times faster than growth was. Understanding why that switch happens is the key to understanding everything else about microtubule biology.

How growth works at the plus end

Growth happens because free tubulin dimers in the cytoplasm carry GTP bound to the beta-tubulin subunit. When a GTP-tubulin dimer lands on the plus end, it fits snugly onto the existing lattice and locks into place through lateral contacts with neighboring protofilaments and longitudinal contacts with the dimer below it. GTP-tubulin dimers adopt a relatively straight conformation, which lets them pack tightly into the cylindrical wall. As long as GTP-tubulin keeps arriving faster than hydrolysis can convert it to GDP-tubulin, the plus end stays in a growth-favoring, straight-tipped configuration.

The rate of this addition scales with how much free tubulin is available. Higher tubulin concentration means more frequent dimer arrivals, faster growth, and a more robust stabilizing cap. Lower concentration slows arrivals, thins the cap, and pushes the microtubule toward catastrophe. This concentration dependence is one of the main ways cells and experimenters can tune growth behavior.

The GTP cap: why it matters and how small it really is

Here is where the mechanism gets genuinely surprising. As each GTP-tubulin dimer is incorporated into the wall, it eventually hydrolyzes its GTP to GDP. That hydrolysis event changes the dimer's preferred shape: GDP-tubulin wants to curve outward, away from the tube axis. Inside the lattice, surrounded by neighbors on all sides, it is forced to stay straight, but the stored mechanical stress makes the lattice intrinsically unstable. The GTP-tubulin at the very tip has not yet hydrolyzed, so it sits on top of this strained GDP lattice like a cap on a pressurized bottle.

How thick is that cap? Classic dilution experiments from the early 1990s gave a striking answer: very thin. When researchers suddenly washed away free tubulin to stop new additions, microtubules typically underwent catastrophe within about one second. That short delay implies the cap contains only a few layers of GTP-tubulin, not dozens. The cap size appears to be essentially independent of how fast the microtubule was growing just before dilution, which means that faster growth does not automatically equal a thicker, safer cap.

How shrinkage happens: catastrophe explained

Catastrophe is the abrupt transition from growth to rapid shrinkage. It happens when the GTP cap is lost, either because hydrolysis catches up with the growing tip or because new additions stop arriving fast enough to maintain it. Once the cap is gone, the strained GDP-tubulin at the tip is exposed. The curved GDP-tubulin protofilaments peel outward like the petals of a blooming flower, breaking the lateral bonds that hold the tube together, and the whole end rapidly unzips. Subunits fly off at a rate many times faster than they were added during growth.

Because GTP hydrolysis is stochastic, catastrophe is also stochastic. A given microtubule does not catastrophize at a predictable moment; instead, it has a certain probability per unit time of losing its cap. That probability is what experimenters call the catastrophe frequency, and it is measured as the number of growth-to-shrinkage transitions divided by the total time spent in growth.

One nuance worth knowing: catastrophe frequency during disassembly can behave somewhat independently of bulk free GTP-tubulin concentration under some conditions. This tells you that it is really the structural and nucleotide state of the tip itself, not just the solution around it, that drives the transition. The lattice history matters.

Rescue: switching back to growth

Rescue is the opposite event: a shrinking microtubule suddenly stops shortening and starts growing again. It is rarer than catastrophe under pure in vitro conditions, but cells actively promote it. Rescue frequency is measured as the number of shrinkage-to-growth transitions divided by the time spent shrinking. One model proposes that rescue happens when a shrinking end encounters a region of lattice that still retains some GTP-tubulin or GDP-Pi (the intermediate state between GTP and GDP), which provides just enough stability to pause the peeling and allow new GTP-tubulin to re-stabilize the tip. Lattice defects and damage-repair events may also seed rescue events inside cells.

Proteins that bias growth, catastrophe, and rescue

Cells do not leave dynamic instability to pure chance. A whole toolkit of microtubule-associated proteins (MAPs) and motor proteins push the system toward growth, catastrophe, or rescue depending on what the cell needs at that moment.

XMAP215: the growth accelerator that also promotes catastrophe

XMAP215 (and its family members, including Dis1 in yeast and ch-TOG in humans) acts like a polymerase at the plus end, delivering tubulin dimers to the tip and dramatically accelerating growth rate. In reconstitution experiments, adding XMAP215 can increase growth velocity by 5 to 10 fold. Counterintuitively, it also raises catastrophe frequency. The working model is that XMAP215 speeds up both addition and the structural maturation of the end, so the cap transitions through its states faster, increasing the probability of stochastic cap loss. In cells, XMAP215 partners with EB1 at growing plus ends, and the two together create the characteristic comet tail visible in fluorescence imaging.

CLASPs: the rescue promoters

CLASP proteins are the main anti-catastrophe, pro-rescue regulators. In vitro, adding CLASP decreases catastrophe frequency and increases rescue frequency, and it can do this without dramatically changing growth or shrinkage velocity. The mechanism involves CLASP's TOG domains recruiting tubulin dimers to the tip of a shrinking microtubule, effectively re-seeding a new cap. This makes CLASPs particularly important at places in the cell where microtubules need to survive near the cortex or at kinetochores, because a microtubule rescued from catastrophe does not have to be nucleated all over again from scratch.

Kinesin-13: the active depolymerizer

Not all catastrophe is passive. Kinesin-13 proteins (like MCAK in vertebrates) use ATP hydrolysis to actively peel GDP-tubulin protofilaments from microtubule ends, driving rapid depolymerization without needing to wait for spontaneous cap loss. Cryo-EM work shows kinesin-13 engages the curved, peeling-like protofilament conformations that characterize the shrinking end, essentially amplifying the structural instability already present in GDP-tubulin. Cells deploy kinesin-13 at kinetochores and throughout the cytoplasm to tune the catastrophe rate on demand.

EB proteins and plus-end tracking

EB1 (and EB3) are perhaps the most useful markers of growing microtubule ends because they specifically recognize the structural and nucleotide state of the GTP-cap region. EB1 binds most strongly to a transitional lattice state that is neither pure GTP nor pure GDP, meaning EB comets actually report structural maturation of the cap rather than simply GTP presence. This is practically important: EB comet length and intensity in live cells can be interpreted as a readout of how fast the cap is maturing and how stable the growing end is. EB1 also recruits other plus-end tracking proteins (+TIPs) including CLIP-170, meaning that the entire plus-end regulatory network depends on EB1 as an organizing scaffold.

What controls whether growth or shrinkage wins

Four categories of conditions set the balance between growth and catastrophe. Understanding them helps you predict how changing one variable will shift microtubule behavior.

The nucleotide state is the most direct driver. GTP-tubulin is straight and sticky; GDP-tubulin is curved and wants to peel. As these ends repeatedly gain and lose the nucleotide and structural states that drive peeling, they weave webs as they grow GTP-tubulin is straight and sticky; GDP-tubulin is curved and wants to peel. The ratio of GTP-tubulin to GDP-tubulin in the pool sets how often stabilizing subunits arrive versus how often the cap can be breached. Cells maintain a large pool of soluble GTP-tubulin partly for this reason: it keeps microtubules in a growth-competent state by default, and active regulation then decides when catastrophe is triggered.

Ionic conditions, especially Mg2+ and total ion concentration, matter more than is often appreciated. Divalent cations influence the geometry of GTP binding and the rate of hydrolysis, meaning that in vitro results can shift noticeably between labs simply because of buffer composition differences. This is one reason why reconstitution experiments need careful standardization to be comparable across labs.

How researchers actually measure all of this

In vitro TIRF reconstitution assays

The standard lab approach is Total Internal Reflection Fluorescence (TIRF) microscopy using a reconstituted system. Short GMPCPP-stabilized microtubule seeds (GMPCPP is a slowly hydrolyzable GTP analog that keeps the seeds stable) are immobilized on passivated glass. You then flow in fluorescently labeled soluble tubulin plus GTP, and watch dynamic extensions grow from the seed ends in real time. The system is clean enough that you can add back individual MAPs one at a time and measure exactly how each one changes the dynamics.

From those movies, you extract four core parameters using kymographs (space-time plots where the x-axis is position along the microtubule and the y-axis is time). The standard definitions are:

1. Growth velocity (vg): how fast the plus end extends, in micrometers per minute
2. Shrinkage velocity (vs): how fast the end depolymerizes during catastrophe
3. Catastrophe frequency (fc): number of growth-to-shrinkage transitions divided by total time spent growing
4. Rescue frequency (fs): number of shrinkage-to-growth transitions divided by total time spent shrinking before reaching the seed

Washout (dilution) experiments are a classic add-on to TIRF assays. Once a microtubule is growing, you abruptly flush away free tubulin. The microtubule keeps growing for a brief moment while the existing cap decays, then undergoes catastrophe. The distribution of that delay time tells you about cap size and decay kinetics. Short mean delays (around 1 second) are direct evidence that the stabilizing cap is thin.

Live-cell imaging

In cells, the most informative approach is to fluorescently tag EB1 or EB3 and watch the growing comets in real time. Because EB1 specifically binds the GTP-cap-associated structural state, comet length and brightness are a proxy for cap health. Longer, brighter comets suggest a more stable growing end; short or flickering comets precede catastrophe. Combining EB1 imaging with tracking of a second marker (like CLIP-170 or XMAP215) can reveal how the plus-end protein network responds to changes in tubulin availability or MAP activity.

High-resolution nanoscale imaging of EB1 distributions in interphase cells has shown that the comet shape reflects maturation transitions happening at the growing tip on the timescale of seconds. Faster maturation correlates with shorter cap lifetime and higher catastrophe probability, giving you a mechanistic readout from a simple fluorescent marker. If you want to investigate how a specific MAP or drug shifts microtubule behavior in living cells, measuring changes in EB1 comet length, density, and dynamics is often the most direct readout available.

Connecting microtubule dynamics to broader growth logic

Dynamic instability is not just a molecular curiosity. It is the physical mechanism that lets cells search space efficiently, build the mitotic spindle, and position organelles. A microtubule that only grew would be rigid and eventually trapped. A microtubule that only shrank would disappear. The switching behavior, biased by MAPs and tubulin concentration, lets a cell probe its interior rapidly and then commit resources to whichever direction finds a target. It is the same logic as adaptive growth in any biological system: explore, then exploit.

Tubulin concentration sets a hard physical limit on how fast and how stably microtubules can grow, just as nutrient availability limits cell growth or water availability limits plant elongation. When tubulin is scarce, the GTP cap thins, catastrophe dominates, and the microtubule network collapses. This is why dividing cells upregulate tubulin expression before mitosis and why drugs like taxol (which stabilizes the GDP lattice) or vinblastine (which sequesters tubulin) can so effectively disrupt the spindle: they shift the growth-catastrophe balance far from the narrow window where useful dynamics occur. The same principle applies when thinking about how the plus end differs from the minus end, or why certain cell types with unusually long microtubules (like neurons) require especially robust CLASP and XMAP215 activity to keep those structures from catastrophizing before they reach their targets.

If you are exploring related aspects of cellular growth and dynamics, the mechanics of how microtubules grow from their two ends and whether both ends contribute to elongation are closely tied to everything discussed here. The end-specific behaviors described in this article also connect to how plant cells orient their cortical arrays and guide directional growth, since the same dynamic instability rules govern how plant microtubules organize themselves in response to developmental cues. In pollen tubes, the same type of tip-focused growth is promoted by signals that stimulate the cytoskeleton to keep polymerizing at the front, which helps answer what stimulates the pollen tube to grow plant cells orient their cortical arrays and guide directional growth. In plant systems, that same dynamic instability framework lets researchers predict how the coleoptile will grow over time.